I need to take pains to distinguish my thinking from the so-called Formalistic reasoning of the early 20th century (see Brian Tamanaha’s Beyond the Formalist-Realist Divide for some good reading). I don’t think the answer to legal disputes can be scientifically deduced from the precedents (this was the position Holmes rejected). Rather, based on legal precedents, one can predict what a given judge/court will do with a specific legal problem. This is rather legal realist. Based on what a judge did before, he will do again. It’s not the precedent that guides him, by itself, but how he views that precedent.